{
  "id": 28396,
  "title": "Machine learning and online tests identify autistic adults with 92% accuracy",
  "url": "https://urgent.news/2026/08/01/machine-learning-and-online-tests-identify-autistic-adults-with-92",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-01T20:00:31.000Z",
  "source": {
    "name": "PsyPost",
    "slug": "psypost",
    "url": "https://www.psypost.org/machine-learning-and-online-tests-identify-autistic-adults-with-92-accuracy/"
  },
  "original_language": "en",
  "account": "A groundbreaking study published in Translational Psychiatry has demonstrated that a combination of machine learning algorithms and online cognitive tests could accurately identify adult autism with a remarkable 92% accuracy. This research, led by Erik Van der Burg and Robert M. Jertberg from Vrije Universiteit Amsterdam, analyzed data from 552 Dutch adults, with 286 diagnosed with autism and 266 without. The participants completed various online tasks focusing on emotion recognition, perception, and mental skills, generating 54 measurements. Initially, the model achieved 81.8% accuracy in the full sample, but after accounting for age and gender, the accuracy increased to 74%. When combined with the autism questionnaire, the overall accuracy reached 92%, successfully identifying 88% of autistic participants and 96% of non-autistic participants. Reaction times during emotion recognition tasks proved particularly informative for the model. The researchers discovered that while individual test scores often lacked significant differences, an overall pattern across several tasks could accurately predict autism diagnosis. However, the study has limitations, including a sample population with generally higher intelligence and fewer support needs than the broader autistic community, and a comparison to the general public rather than clinical referrals. Nonetheless, this innovative approach shows promise in overcoming the difficulties in diagnosing autism in adulthood, where lengthy waiting lists often hinder access to proper diagnosis and support.",
  "summary": "Diagnosing autism in adults can be difficult due to long wait times. A new study shows that combining a standard questionnaire with online mental tests and artificial intelligence can predict an autism diagnosis with 92% accuracy.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 2,
    "also_reported_by": [
      {
        "outlet": "bioRxiv",
        "title": "Combining Machine Learning and Directed Evolution for Optimization of a Monooxygenase",
        "url": "https://urgent.news/2026/08/03/combining-machine-learning-and-directed-evolution-for-optimization-of",
        "published": "2026-08-03T00:00:00.000Z"
      }
    ]
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}